Review by: Park, Jonghwi (2005-11-22)

This is a controversial paper that has provoked, since its publication in 1983, a great deal of lively and continuing debate in the field of instructional media research.

The author claims that different media do not influence learning because they are merely vehicles for the delivery of instruction just as the trucks that deliver our food do not influence our nutritional states; therefore, it is meaningless to ask whether one form of medium is more beneficial than another form when it comes to learning. The author first supports his claim with his review of some meta-analyses of media research. Then, he discusses the more recent “media attribute” theories and “symbol system” theories. Finally, he suggests directions for future research on media.

His review of meta-analyses of media research indicates that media comparison studies that find causal connections between media and achievement are confounded for two reasons:

1. Effect sizes of different treatments significantly decrease when instructional methods and contents are controlled. Thus, it is not media per se but uncontrolled instructional methods and contents that lead to differences in learning achievement.
2. Uncontrolled novelty effects of newer media have been misreported as significant learning effects. For example, in a review of computer-assisted instruction, computer effects were .56 standard deviations for studies lasting 4 weeks or less but were reduced to .2 standard deviations in studies lasting 8 weeks.

The author concludes that five decades of media research shows that there are no learning benefits to be gained by employing one form media rather than another. All learning benefits reported in media comparison studies are the result of confounding media with methods of instruction.

In the second part, the author examines more recent directions of research that investigate “attributes” of media, symbol systems, and their influence on the cultivation of cognitive skills. He argues that certain attributes of media such as zooming or animated arrows might serve as “sufficient” conditions for facilitating learning; however, instructional theories seek “necessary” conditions where the instructional method models the cognitive processes required for the successful performance of a learning task. He presents research indicating that the necessary condition for acquiring cue-attending skills in a visual field was the efficient isolation of relevant cues rather than zooming or irising. He suggests that the determination of necessary conditions is a more fruitful approach since once described, the necessary cognitive operation becomes a recipe for an instructional method.The author then suggests that individual beliefs and preferences about media are promising directions of further research involving media since these learner attributions are causally connected with the efforts they invest in learning.

In conclusion, the author asserts that there are no learning benefits to be gained from media, only economic benefits from media. Consequently, future research should focus on the “necessary” characteristics of instructional methods and other variables (task, learner aptitude, and attributions), which are critical to understand how learning occurs and how achievement increases.

Although the author made his argument based on the existing evidence derived from media research in the past, this paper created an enormous amount of controversy in media research. The author’s response to the debate, “Media Will Never Influence Learning” might help readers to understand how the debate unfolded and how he has updated and sharpened his argument.